<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="lee-etal-2026-latent">
    <titleInfo>
        <title>Latent Task Embedding: Geometric Expansion for Single-Path Multi-Task PEFT</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Junhyun</namePart>
        <namePart type="family">Lee</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Seoha</namePart>
        <namePart type="family">Song</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Dhananjaya</namePart>
        <namePart type="family">Gowda</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-10</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Hung-Yi</namePart>
            <namePart type="family">Lee</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Sunipa</namePart>
            <namePart type="family">Dev</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jessy</namePart>
            <namePart type="family">Li</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Desmond</namePart>
            <namePart type="family">Elliott</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Budapest, Hungary</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-459-0</identifier>
    </relatedItem>
    <abstract>Parameter-efficient fine-tuning (PEFT) typically relies on additive weight perturbation. However, in dynamic deployment environments like multi-tenant cloud serving or mobile edge devices, additive methods force a dilemma: merge weights and lose task-switching flexibility, or remain unmerged and suffer severe parallel-branching latency. To break this bottleneck, we propose Latent Task Embedding (LatTE), which shifts the PEFT design axis to geometric expansion. Instead of perturbing frozen parameters, LatTE concatenates a task-specific auxiliary embedding to the input and expands the corresponding weight matrices. This structural fusion ensures the forward pass remains a single, optimized matrix multiplication, bypassing the routing overhead of unmerged adapters. Furthermore, LatTE natively supports router-free, intra-batch multi-tasking by isolating tasks into distinct subspaces controlled via binary masking. Extensive evaluations across 1B to 14B parameter models demonstrate that LatTE matches highly optimized baselines like LoRA, while raising throughput by 1.7× over routed multi-adapter LoRA under heterogeneous-task batching.</abstract>
    <identifier type="citekey">lee-etal-2026-latent</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.2649/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>60275</start>
            <end>60293</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Latent Task Embedding: Geometric Expansion for Single-Path Multi-Task PEFT
%A Lee, Junhyun
%A Song, Seoha
%A Gowda, Dhananjaya
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F lee-etal-2026-latent
%X Parameter-efficient fine-tuning (PEFT) typically relies on additive weight perturbation. However, in dynamic deployment environments like multi-tenant cloud serving or mobile edge devices, additive methods force a dilemma: merge weights and lose task-switching flexibility, or remain unmerged and suffer severe parallel-branching latency. To break this bottleneck, we propose Latent Task Embedding (LatTE), which shifts the PEFT design axis to geometric expansion. Instead of perturbing frozen parameters, LatTE concatenates a task-specific auxiliary embedding to the input and expands the corresponding weight matrices. This structural fusion ensures the forward pass remains a single, optimized matrix multiplication, bypassing the routing overhead of unmerged adapters. Furthermore, LatTE natively supports router-free, intra-batch multi-tasking by isolating tasks into distinct subspaces controlled via binary masking. Extensive evaluations across 1B to 14B parameter models demonstrate that LatTE matches highly optimized baselines like LoRA, while raising throughput by 1.7× over routed multi-adapter LoRA under heterogeneous-task batching.
%U https://aclanthology.org/2026.emnlp-main.2649/
%P 60275-60293

Markdown (Informal)

[Latent Task Embedding: Geometric Expansion for Single-Path Multi-Task PEFT](https://aclanthology.org/2026.emnlp-main.2649/) (Lee et al., EMNLP 2026)

• Latent Task Embedding: Geometric Expansion for Single-Path Multi-Task PEFT (Lee et al., EMNLP 2026)

ACL

• Junhyun Lee, Seoha Song, and Dhananjaya Gowda. 2026. Latent Task Embedding: Geometric Expansion for Single-Path Multi-Task PEFT. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 60275–60293, Budapest, Hungary. Association for Computational Linguistics.